The day before Kintan Brahmbhatt’s 10th birthday, he stayed up late, gripped with excitement, so he could open his presents. After tearing off the wrapping on one package, he was overjoyed to find a bright red, remote-controlled toy car inside. Brahmbhatt placed the car on the floor. With a smile on his face, he inched the joystick forward.
The race car didn’t move.
As it turned out, the toy didn’t come with batteries, nor were there any in the house. At that hour, there was no way to get them. Where Brahmbhatt lived, there were no stores open late. Deflated and disappointed, he went to bed.
This was Brahmbhatt’s earliest brush with what he calls friction — anything that gets in the way of a customer and a task. Put another way, it’s any obstacle that prevents a user from trying or using a product or service. Since the failed spin on the race car, Brahmbhatt’s built a career based on taking friction out of technology, often in remarkably innovative ways. At Amazon, he’s helped develop and define the product strategy behind Amazon Music subscription services and music on Alexa, the voice-recognition technology behind the company’s smart speaker, Echo. He also previously headed the company’s IMDb Everywhere initiative, which produced the “X-Ray” feature that automatically clues viewers into the names of the actors in any given scene. Before that, he helped build what became Skype for Business at Microsoft.
In this exclusive interview, Brahmbhatt takes us through how to detect and anticipate points of friction through monitoring steps in a customer’s journey with your product. He shares the three stages of the product experience where customers are most vulnerable to experiencing friction. Brahmbhatt also walks through strategies to reduce friction, or, if it’s unavoidable, how to mask it.
Friction can manifest in a number of ways, ranging from overly complicated sign-up procedures to difficult-to-navigate menus. According to Brahmbhatt, there are three types of friction:
Friction due to unfamiliarity. A customer’s first-time experience with your product automatically contains friction when they don’t know how to use it. This is one of the biggest hurdles to get over when launching a new product.
Friction by design. There are times when you have to intentionally build friction into a product. If your product has a learning curve, you’ll have to design that experience thoughtfully.
Friction due to misalignment with human behavior. Here’s where design can come back to bite you. Poorly placed app controls, failing to anticipate how consumers will use the product and badly designed navigational tools will cause this.
While the first two types of friction are easier to anticipate and control, the third category includes cases that are unintentional or emerge unexpectedly. “I got a mega-shock when we launched music services on Alexa. Customers were trying to find music in ways none of us had ever imagined. We’d been prepared for people making specific song requests,” Brahmbhatt says. “But the problem with new releases is that people don’t necessarily know the name of every song. The voice commands that customers asked Alexa were more conversational than we’d expected. It was, ‘Play the latest single by Adele’ or ‘Play the new album by Adele.’ Or for Sia’s ‘Chandelier,’ they’d ask for it by the lyric — ‘Play one-two-three by Sia.’”
As a product line — or a music service — evolves, more points of friction can arise. “Customers would say ‘skip’ after asking for a U2 song — Alexa would cue up the band’s latest tune, when the listener really wanted to hear music from the 1980s. We were missing the point in addressing the customer’s intent. There was so much innate friction in the customer’s ability to get the music they wanted to get,” Brahmbhatt says. “The goal here is really about anticipating those friction points in the customer’s mind — and if you can’t, then finding and alleviating them quickly. This isn’t easy. You have to study failure points — for us, it was the number of customer requests that were not fulfilled, or the number of songs cued up that were skipped.”
Despite best efforts to anticipate and address friction, know that there will be times you don’t find out about a point of friction until it’s too late for a customer. In Brahmbhatt’s early days working on digital medical records project Securamed, an early user traveled to Mexico on vacation and found herself in need of medical attention. Unfortunately, Securamed stored records in English only, and her Spanish-speaking doctor was unable to use them. It was an example of unintentional friction — which you frequently see in new products. The team later added multi-lingual capabilities, and use of the product climbed as a SARS outbreak hit Asia.
When building products, you’re always either removing, adding or masking friction.
So how do you catch friction before it manifests? First, know where to find it. This is where knowing your customer’s journey comes in. You’ll want to hone in on the specifics, but for any product, there are three stages to how a customer interacts with your company’s offering:
Before first contact: This is the easiest one to spot. “At the very beginning of your product’s life, the customer isn’t aware it even exists. Your goal is to convert them from being unaware to being interested. It’s all about anything you can do to remove the friction that’s preventing your ideal customer from knowing you’re out there,” Brahmbhatt says. “You can measure awareness by looking at your market segment share and other leading indicators, like press and social media mentions. When the press is writing about your industry, are they mentioning your product or service? Look at web and app store traffic, installs, number of reviews and the quality and sentiment of the reviews. Work with your marketing team on improving these metrics.”
Sign up + one “transition” task. Once someone is aware of your product, more potential for friction enters the picture. “As the user moves along in the journey of your product, they go from being interested to being a customer, or at least somebody who’s tried it at least once. You want them to sign up and then perform one meaningful task,” Brahmbhatt says. “So often startups lean on how many signups they’ve had, but if customers don’t engage with it in some meaningful and measurable way, they’re unlikely to return. For example, if it’s a healthcare app, maybe the goal is for the user to sign up and book their first doctor’s appointment. This task sets expectations for a future engagement.”
How to choose your “transition” task: “You don’t want it to be too complex. Focus on the behavior that you want the customer to repeat.” Brahmbhatt says. “Pick the action that helps a user repeat the task over and over again. In the case of a healthcare app, it’s booking an appointment or asking a question to the physician. In banking, it’s signing in to check your balance for the first time. In music, it’s about playing the first song — so that you can get suggestions for the next song.”
First moment of delight. Now that people are engaged, this is your opportunity to impress them. “This is where you transform someone from being an occasional user to being a fanatic customer. To add the delight factor, look for counterintuitive patterns in your data. Look for failed use cases that you would have expected to succeed. With Alexa, we saw that customers’ demeanor became more casual when it came to music,” Brahmbhatt says. "Now that people are engaged, this your opportunity to impress them. We made it so that it would play music inspired by what they’d recently listened to. Or in December, it would contextually cue up holiday music. When Alexa responded to these casual voice commands and played music that customers loved, they were delighted and were more likely to come back and use it again.”
Indifference generates the highest friction for new products.
No one would argue with the advice that one should listen to their customers. But when and where can be harder to determine, because ultimately you’re trying to intercept key moments that illustrate human behavior. Once you do, your goal is to conform to it or change it. “First I’ll say that it's very hard to change human behavior. Unless it's absolutely necessary, I try to avoid changing behavior on day one. It can be a gradual journey where it takes days or even months depending on the type of behavior,” Brahmbhatt says. “However, it’s doable. You have to observe existing human behavior and map your products to it.”
According to Brahmbhatt, there are three types of listening posts:
Natural Habitat. Think the Nature Channel. There are a myriad of tools that allow product teams to monitor customers in-app, but don’t neglect getting in the field. That can mean getting invites into offices, retail stores or living rooms — all depending on your target customers. “Shadow your customers. It’s inexpensive and very easily doable by small start-ups. You don't really need to observe 1,000 people — five or so will do,” says Brahmbhatt. “While researching, we’d watch how people read books or how families listen to music. We’d see customers rewind a song by a few seconds to listen to a particular lyric, or using X-Ray to jump to a particular scene to see Alfred Hitchcock’s cameo. These trends were once counterintuitive to us. Observing your customers while they use your product gives you a glimpse into the role your product plays in their life, and there’s a treasure trove of information there.”
Mentions and Reviews. Feedback can be explicitly expressed, like app reviews, social media mentions and customer emails, but it can also be something you tease out of the data. “Examine your usage metrics and dig for atypical trends. Pay particular attention to high frequency actions and recurring patterns. You might find that people gravitate to a certain part of your app, or are repeat buyers of certain products,” Brahmbhatt says. “As we were figuring out how to enable customers to buy something via voice, we realized that some customers may need to look at a picture or read a detailed description before making the final decision. We found out that if a customer had bought a product before, his anxiety about buying it again is reduced dramatically. Now, if I ask Alexa to buy toilet paper, she’ll ask me if I want the same 36-pack of Cottonelle I bought last time. All I need to say is yes, and the product shows up at my doorstep.”
The Industry Standard. Find the expectation that's been set in your customers’ minds for existing products and services. “That bar is the current path of least resistance. It could be the search bar always being on top, or the menu being on the left — that’s what your customer’s muscle memory will gravitate towards. You, of course, also want to know about larger trends industry-wide, so that you can offer a service that’s competitive,” Brahmbhatt says. “Friction plays a pivotal role in the launch of a new product. Customers must perceive the value of your product to be very high, and the corresponding effort to try it out has to be very low. The next best alternative to your customer trying your product is to do nothing at all.”
Always migrate your audience to the path of least resistance.
Think about friction as the simplest way for your user to get where she wants to go. “This is what we’re trained to do as human beings — take the easiest route. Reducing friction is all about creating a path of least resistance,” says Brahmbhatt. “But that doesn’t mean you’re always changing customer behavior. You’re splintering off from it. To do so, ask: what’s the closest path to existing behavior that you can create? How can you get a user to complete a task in a more efficient way, but which still maps with her existing behavior?”
These are the steps Brahmbhatt recommends for cutting friction from your product:
Reduce anxiety. Two factors cause anxiety: decisions and loss. The Amazon Music team faced both when music began to migrate from physical CDs to digital. “Customers were stressed out about choosing whether they should buy a CD or vinyl, or purchase an mp3 instead. It created cognitive overhead and friction in the act of listening to music. People couldn’t decide, and having to added anxiety,” Brahmbhatt said. “There also was the fear of loss. Some of our customers had purchased CDs that were worth hundreds of dollars since 1998. Should they just lose it in the migration to digital? They could rip the CD, but that would take a lot time and effort. So we launched AutoRip, which automatically gave customers MP3 versions of CDs and vinyls they’d purchased. Customers were thrilled, and we eliminated two causes of anxiety —decisions and loss — with one product launch.”
Remove avoidable steps. When designing your product, ask yourself if you need the customer to make a decision. “Make a list of all decisions a customer needs to make in order to complete a task in your product. Then question each one. For example, in a healthcare app, since your home address is already in there, the app could automatically select a clinic based on proximity to your location. That takes the friction and anxiety of choice out of the process,” Brahmbhatt says. “When setting up a music player, a customer could choose the streaming bitrate — a measure of audio quality — of up to 256 kilobytes per second (kbps). We asked ourselves two questions – first, if the customer absolutely needed to make that decision or whether we could pre-select it for the user. To reduce cognitive load, rather than calling it 256 kbps, we called it ‘high.’ We offered options that were easy to understand — low, medium, high and then added a pre-selected choice of ‘auto.’ You as a customer didn’t need to worry about making that decision. It was already taken care of for you.”
Another way to get at avoidable steps is to ask yourself these questions, in this order:
How many decisions are between a customer and completing a task?
Are each of these decisions absolutely necessary?
If so, can you make the decision for the customer by pre-selecting an option?
If not, and the customer absolutely needs to make that decision, how can you simplify the decision process?
If there are multiple decisions, could you combine them into one decision? Can you present the most important decision first to the customer?
How can you preserve the decision once it’s been made so that you don’t have to ask the customer again in the future?
Mitigate context switching. Context switching often happens when a customer must navigate away from your app or site to complete a task. It’s the point at which your customer will abandon your product. “To solve this, create points where customers don’t have to switch contexts. For instance, Buffer and Pocket have Chrome extensions that allow you to operate their services from your web browser, and they do it beautifully. Look at some of the most successful apps out there. The check-in button on Foursquare was always apparent no matter where you navigated in the early versions of the app,” Brahmbhatt says. “Allow customers to easily switch back and forth between the primary task — in Foursquare, that was checking in — and the secondary task, which was browsing restaurants. In early days of Kindle, we noticed that some customers would leave the book to look up words in the dictionary, so the team built that function into the Kindle app. Along similar lines, customers would also navigate away from movies to look up actors on IMDb. We addressed that by allowing them to easily find out the actor on the screen while watching a movie or a TV show on Amazon video. Instead of teaching customers a new UI action, we automatically showed names of the actors on screen whenever the user hit pause while watching the video.”
Sometimes there’s no way to avoid context switching. “If you absolutely have to do it, then do it gracefully. When customers switch between listening to music on their computer versus the Echo at home, we make it easy to recall recently played songs. Audible also syncs seamlessly with the Kindle, so that you can pick up an ebook from the same spot where you left off in the audiobook,” Brahmbhatt says. “The top two issues around context switching are continuity and recall. Customers want to be able to recall what they did earlier — a movie they watched, a song they played. They also want continuity, so letting the customers pick up from where they left off is critical.”
Friction is unavoidable in some cases. “If there’s just no way to avoid it, mask it. When we launched AutoRip, the app would take a few seconds to load the large music collection for customers who had purchased thousands of songs in the past. While it loaded, instead of showing a simple progress bar or a spinner, we masked that moment of pause by greeting the customer with a useful welcome message to explain the service,” Brahmbhatt says.
Friction may be added by design to produce later benefits. “For subscription services, it is often effective to place the friction upfront by requiring the credit card on file so there’s no extra step needed at the end of the trial in order to subscribe. Host verification on Airbnb, or presenting terms and conditions are other examples where friction is introduced by design in the customer journey. A little friction can also lead to increased usage and loyalty. For instance, customers who up or down vote a song on a radio station feel more invested in the station they’ve helped curate,” Brahmbhatt says. “If they’ve done some work to inform their experience, it increases the likelihood of repeat usage. Customers feel they were involved in building something that they enjoy. It’s important to pick the right place to go when you’re adding friction intentionally. Identify the most commonly repeated actions and get new customers to perform the first iteration of those actions. Playlist creation is an example of this. Customers who create a playlist by investing some effort in the music app end up using the app more frequently.”
Add friction to create a sense of ownership — carefully. The goal isn’t for the user to do a task, but to do it in the most satisfying way.
Friction can be a powerful deterrent between a customer and your product. It’s typically introduced by a lack of familiarity or misalignment with human behavior. Start to combat it by mapping the three stages of friction: before the first contact, the signup process — during which you want the customer to complete one task — and the first moment of delight. Then set up listening posts for observing your customers and teasing counterintuitive insights out of the data to discover where friction is occurring. Finally, cut friction by reducing customer anxiety, cutting avoidable steps and preventing context switching.
“A customer’s experience with your product should be smooth and seamless. Start with the customers and work backwards. Listen to and observe your customers closely. Find the potholes and bumps in the road. Where do they get slowed down or tripped up?” Brahmbhatt asks. “Carve out the path of least resistance for your customer. Always stay true to what’ll take friction out of the process — products that falter are often ones where this has been neglected. It takes effort, but it’s worth it. And it’s not always as hard as you might think. Sometimes all that stands between your product and a moment of delight is a pair of batteries.”
Photography by Bonnie Rae Mills.